Wait but actually why: Brain-Machine Interfaces and Unit Economics of Human Output
This article was featured on The Indian Economist and Mattermark Daily’s twitter handle
Last week, Tim Urban wrote a ~36000 word blog post on Wait But Why, on how and why Elon Musk is working on Neuralink to develop the Brain Machine Interface (BMI) of the future. This is the second consecutive (public) development of the week for BMIs, after Facebook also announced in F8 2017 that they are working on a technology that will eventually allow people to type at 100 words per minute, 5X faster than typing on a phone, with just your mind.
Quick perspective for the uninitiated (if you are a bit lazy to go through 36K words) — the advent of Internet of Things allowed machines to communicate with each other and take decisions based on pre-fed algorithms and artificial intelligence. Similarly, in ideal state, Brain-Machine Interfaces will allow human brains to directly communicate with each other and with machines, without the need of voice or actions (‘Internet of Brains’?).
Imagine sending a brain to brain SMS to someone else by just ‘thinking’ — all with the help of a device that sits on top of your head, running on a ‘Brain OS’, and connected to cloud servers processing billions of Brain messages. Surreal.
One of the underlying themes in what both Elon and Zuck are trying to achieve (for two very different reasons, perhaps) is to unlock the realm of human productivity. But why, really? In this post, we look at how BMIs affect productivity by understanding the unit economics of human output.
Unit Economics of Human Output
Ideally speaking, human output at unit level is fundamentally dependent on two factors — available time and number of things that can be achieved by a human in unit time (productivity)
For n=1 human
Daily Human Output (DHO) = total time x f(productivity) — — (1)
Aside: f(productivity) means function of productivity; ‘— — (1)’ indicates equation 1
Many of us have different variables at play wired in our lives for available time and productivity, and desired output is also very subjective from person to person.
Part A of DHO: total time = time at work + time at leisure + time at sleep — — (2)
Total time in a day is constant . We can have more ‘available time’ at our disposal by sleeping for only 2 hours everyday, but in ideal cases, our productivity function f(productivity) will likely take a hit (even after gulping down multiple cups of coffee) and our DHO will go down.
However, let’s assume you sleep as much as the National Sleep Foundation recommends — about 7–9 hours a day if you are 18–64 years while reading this — that means your daily human output is now significantly dependent on what you can make of the time you are awake, aka productivity, since the incremental effect from altering time at sleep is either too low a benefit (~1–2 hours everyday) or has a negative effect on f(productivity).
Part B of DHO: f(productivity) = IQ x EQ x f(others) — — (3)
Your productivity will depend on how intellectually and emotionally intelligent you are to tackle a specific situation. ‘f(others)’ accounts for cases such as availability of gadgets that can increase productivity in certain situations — like how you are able to compute 29120 x 123 with the use of a calculator in seconds irrespective of your original IQ and EQ.
Assuming total time in a day to be constant as per current scenarios, increasing human productivity offers a far greater upside than say, countering sleep.
So when Netflix CEO said his real competitor is sleep, I feel he missed the point that Elon and Zuckerberg did not.
For Elon, who claims to sleep 6 hours everyday, will now be running three companies as CEO and has been consistently vocal about the apocalyptic effects of AI-run-machines becoming more intelligent than humans, increasing human productivity/intelligence has a lot of dollar and happiness value.
For Zuck and Facebook, making you spend more time on Facebook networks is one objective and making you spend time more efficiently is another objective — both of these have a lot of dollar and happiness value for Zuck. Let’s dig deeper to know how:
For n = 1 user
Facebook Revenue ∝ Number of impressions ∝ (Time Spent per session x Productivity factor of each session x number of sessions) — — (4)
Facebook Revenue can be increased in many ways, such as
Making time at work (refer 2)= time on Facebook (Facebook Workplaces), which will increase number of sessions
More importantly in current context, by increasing the productivity of your each session on Facebook. If you are making X interactions in the ~1 hour average time you spend on Facebook everyday and Zuckerberg can make that say, 5X by altering ‘f(others)’ (refer 3), Facebook gets to show you 5 times more number of ads in the same time and hence increase ARPU.
There is tangible merit for these folks in making humans more productive.
f(productivity) & Brain-machine Interfaces
So how do humans become more productive? The answer that Elon and Zuck want us to believe is — Brain-Machine Interfaces. By current technology standards, we can not have more than 24 hours in a day, time travel is still fiction and sleep will likely have certain share of that time each day (Damn it, brain). But with BMIs, each factor in the f(productivity) function will be transformed forever.
f(productivity with BMIs) = (B1 X IQ )x (B2 x EQ) x b(others) — — (5)
Aside: B1 and B2 represent multiplication factors for IQ and EQ and b(others) represents modified function f(others) due to BMIs
Intelligence Quotient: Our IQ as an individual remains almost constant in our lifetime (however, our collective IQ at a generation level increases over time), but among many other things, BMIs can help us accelerate our learning curve by many-folds (net output dependent on factor B1 and of course, your IQ). For example, you will be able to learn a skill such as how to code websites (if websites do not get obsolete by then) or how to make a bomb (sadly, bad guys will still exist in the post-BMI world) by ‘downloading’ it directly to your brain and cutting the need to ‘see’ and ‘read’ it in real — oversimplification but well possible (however, you will still need to ‘go through it’ though but much faster speeds should be possible). This also means that all situations remaining same, we will learn more in 1 year in post-BMI world (that number should keep on increasing each year) than we could possibly do next year
Emotional Quotient: Understanding humans is definitely far more challenging and a rare skill-set compared to solving a math equation. HBR writes that the rise of AI makes emotional intelligence more important. BMIs can potentially command control over how you feel or how you perceive things, as well as generate feedback in real time. As an example, BMIs can potentially use AI to tell you what words to say and avoid in a press conference in real time when you are the President of United States (assuming no one is ‘wiretapping’ aka hacking into your Brain OS, in which case they can ‘read your mind’)
b(others): Think brain app store — as an example, what all is possible when the calculator brain app computes and feeds the answer in your brain as soon as you think of the problem.
Summing up:
Daily Human Output (DHO) with BMIs = total time x f(productivity with BMIs) — — (6)
Using 3 and 5,
[f(productivity with BMIs) / f(productivity)] = B1 x B2 x [ b(others) / f(others) ] — — (7)
Using 1, 6 and 7,
[DHO with BMIs / DHO] = B1 x B2 x [ b(others) / f(others) ] — — (8)
Bottomline
Clearly, the impact of Brain-Machine Interfaces on human output depends on
B1: Ability of BMIs to affect our IQ related outputs, such as increasing knowledge, organizing memory or understanding a problem better.
B2: Ability of BMIs to affect our EQ related outputs, such as dealing with mood disorders, understanding emotions, deciphering situations and gaining perspectives.
[b(others)/f(others)]: The factor by which all physical gadgets such as your phone (and all the apps), laptop (and all the apps), calculators etc. transition to being a part of ‘Brain OS’.
2. The lines between AR, VR and AI should blur even further. These mediums should merge in the post-BMI world — think Black Mirror’s Nosedive but without a handheld device — just your brain waves would be enough.
Mixed Reality in the post-BMI world = Augmented Reality + Virtual Reality + Artificial Intelligence + Brain-machine Interface + ______
3. Flip side: All of this sounds exciting, but Tim Urban, in his 36k+ thorough word long post, just scratched the surface of the many negative ways that BMI can impact the world.
This is how this graph should actually look like — which is again, true for any technological innovation.
Considering this post is not meant to deep dive in pros and cons of BMIs, I will skip discussing this further here.
Nevertheless, these are exciting times for technology. The thing about optimistically believing in these prospects is that the Adrenalin rush generated through imagining what all is possible through BMIs to ‘Human Colossus’ shadows the technological challenges of achieving it — typical dreamer vs. realist dilemma. Time and again, history holds it true that the dreamers have consistently expanded what the realist think is real and achievable. Sooner and later, we will get there. We can only hope that this leads to a net increase in gross human output.
Views are personal.
Tweet to me on @videt_jaiswal or reach out on email here
